1. Volume 39, Issue 1, Pages 71-85 (July 2010)
Negative Regulation of Hypoxic Responses via Induced Reptin Methylation 
Jason S. Lee, Yunho Kim, Ik Soo Kim, Bogyou Kim, Hee June Choi, Ji Min Lee, Hi-Jai R. Shin, Jung Hwa Kim, Ji-Young Kim, Sang- Beom Seo, Ho Lee, Olivier Binda, Or Gozani, Gregg L. Semenza, Minhyung Kim, Keun Il Kim, Daehee Hwang, Sung Hee Baek 
Molecular Cell 
Volume 39, Issue 1, Pages (July 2010)
DOI: /j.molcel
Copyright © 2010 Elsevier Inc. Terms and Conditions

2. Figure 1 Reptin Methylation by G9a Methyltransferase
(A) Reptin-interacting proteins were purified from HEK293 cells stably expressing Flag-tagged Reptin by coimmunoprecipitation using anti-FLAG antibody. As a negative control, cells expressing empty vector (Mock) were used. The bound proteins were resolved by SDS-PAGE and prepared for LC-MS/MS analysis.
(B) G9a and Suv39h1, along with previously reported Reptin-binding partners such as HDAC1 and Pontin, were detected from the eluates after purification by immunoblot assays.
(C and D) Coimmunoprecipitation of endogenous Reptin with G9a (C) or Suv39h1 (D).
(E) HEK293 cells were transfected with empty vector or with expression vector encoding G9a or Suv39h1. Cell lysates were immunoprecipitated with antibody against methylated lysine followed by immunoblot analysis with anti-Reptin antibody to detect methylated Reptin.
(F) Autoradiogram of in vitro histone methyltransferase assay using recombinant His-tagged G9a SET domain with either GST-tagged Reptin or Histone H3. Levels of GST proteins used in the assay are shown by immunoblotting with anti-GST antibody.
(G) Immunoprecipitation of methylated Reptin with anti-methyl-lysine antibody from HEK293 cell lysates expressing G9a WT or G9a mutant, followed by immunoblot analysis with anti-Reptin antibody to detect methylated Reptin.
Molecular Cell  , 71-85DOI: ( /j.molcel )
Copyright © 2010 Elsevier Inc. Terms and Conditions

3. Figure 2 G9a Mediates Hypoxia-Induced Reptin K67 Methylation
(A) HEK293 cells were cotransfected with plasmids encoding each GFP-tagged Reptin deletion construct spanning the indicated amino acid residues with or without plasmid encoding G9a. Whole-cell extracts in the presence or absence of G9a were immunoprecipitated with anti-methyl-lysine antibody followed by immunoblot assay using anti-GFP antibody to detect Reptin.
(B) The four lysine residues (K9, K67, K83, and K115) within the N-terminal 135 amino acids were individually mutated to alanine.
(C) Mass spectrometry analysis of Reptin WT and K67A following methyltransferase assay using G9a.
(D) In vitro methyltransferase assay using purified Reptin and G9a proteins using 3H-SAM as a methyl donor.
(E) In vitro methyltransferase assay using Histone H3, Reptin WT, and K67A peptides with purified GST-G9a using 3H-SAM as a methyl donor. Values are expressed as mean ± SEM of three independent experiments.
(F) Dot blotting of unmodified or methylated Reptin peptides with anti-Reptin K67 methyl antibody.
(G) Immunoprecipitation of methylated Reptin with anti-Reptin K67 methyl antibody from cells treated with BIX 01294, a G9a inhibitor, at doses indicated.
(H) Immunoprecipitation of methylated Reptin with anti-Reptin K67 methyl antibody from MCF7 cells exposed to hypoxia for indicated times followed by immunoblotting with anti-Reptin antibody.
(I) Proportion of methylated Reptin was estimated by immunoblotting with either Reptin K67me antibody or Reptin antibody from Reptin immunoprecipitate exposed to hypoxia as indicated.
(J) Reptin methylation was compared between WT and G9a-deficient ES cells in the presence or absence of hypoxic challenge for 9 hr.
(K) G9a−/− ES cells were reconstituted with either WT or enzymatically inactive G9a mutant, and changes in Reptin methylation were monitored in cells exposed to hypoxia with anti-Reptin K67 methyl antibody.
(L) Reptin methylation was determined in MCF7 cells expressing either Reptin WT or K67A mutant. Immunoprecipitation of methylated Reptin was performed as in (H).
See also Figure S1.
Molecular Cell  , 71-85DOI: ( /j.molcel )
Copyright © 2010 Elsevier Inc. Terms and Conditions

4. Figure 3 Reptin-Dependent Target Identification by Microarray Analysis
(A) Flow chart showing the strategy of cDNA microarray analysis and Reptin-dependent gene selection process.
(B and C) Identification of Reptin-dependent (B) and Reptin-independent (C) genes by hierarchical clustering and comparing fold change of hypoxia-induced genes from cells expressing shNS and shReptin. Upregulated and downregulated gene clusters are represented as red and green, respectively. Known target genes for HIF-1, NFκB, TCF/LEF, and p53 were collected, and the enrichment of these genes analyzed for each cluster is shown as a percentage of the total number of genes within each cluster.
(D) Reptin-dependent genes represent about 24.6% of all differentially expressed genes by hypoxia. (E) Heatmap diagram of Reptin-dependent genes (cluster 1) showing gene names with known HIF-1α target genes marked with an asterisk.
(F) Functional classification of Reptin-dependent genes (cluster 1).
See also Figure S2, Table S1, and Table S2.
Molecular Cell  , 71-85DOI: ( /j.molcel )
Copyright © 2010 Elsevier Inc. Terms and Conditions

5. Figure 4
Reptin Methylation Negatively Regulates a Subset of Hypoxia Target Genes
(A) Effect of Reptin knockdown on 3 × HRE-luciferase reporter under normoxia or hypoxia for 9 hr. Results are expressed as fold activation compared to nonspecific shRNA (shNS) under normoxic condition. Efficacy of Reptin knockdown by Reptin shRNA (shReptin) is shown by immunoblot analysis. Values are expressed as mean ± SEM of three independent experiments.
(B) 3 × HRE reporter assay as in (A) with overexpression with Reptin WT and K67A mutant under normoxic and hypoxic conditions. Values are expressed as mean ± SEM of three independent experiments.
(C and D) Quantitative RT-PCR analysis of Reptin-dependent (C) and -independent (D) hypoxia target gene expressions identified following shRNA-mediated knockdown of Reptin and reconstitution of either shRNA-resistant Reptin WT (WTR) or K67A mutant (K67AR) in MCF7 cells in normoxic and 18 hr of hypoxic conditions. Results are expressed as relative mRNA levels compared to shNS under normoxic condition. Values are expressed as mean ± SEM of three independent experiments.
See also Figure S3 and Table S3.
Molecular Cell  , 71-85DOI: ( /j.molcel )
Copyright © 2010 Elsevier Inc. Terms and Conditions

6. Figure 5
Recruitment of Methylated Reptin to Its Target Promoters via Enhanced Binding to HIF-1α
(A) Endogenous interaction between Reptin and HIF-1α in MCF7 cells exposed to hypoxia using either anti-HIF-1α or normal IgG.
(B and C) MCF7 cells were transfected with the indicated expression plasmids, and immunoprecipitation was performed using either anti-HIF-1α (B) or anti-HIF-1β (C) and immunoblotted using antibodies indicated.
(D) Interactions between in vitro-translated 35S-HIF-1α and biotinylated Sepharose beads preincubated with nonmethylated (K67me0) and monomethylated (K67me1) Reptin peptides were analyzed by autoradiography.
(E) HIF-1α-dependent recruitment of methylated Reptin on VEGF promoter in WT and HIF-1α-deficient (HIF-1α−/−) mouse embryonic fibroblasts (MEFs).
(F and G) Transcript levels of Reptin-dependent (F) and Reptin-independent (G) genes with Reptin knockdown in hypoxic conditions. Values are expressed as mean ± SEM of three independent experiments.
(H) The shRNA-coupled ChIP assay on the VEGF and PGK1 promoters in MCF7 cells with time course of hypoxic stress. Promoter occupancy by methylated Reptin, Reptin, HDAC1, HIF-1α, and RNA polymerase II was analyzed.
(I) The shRNA-coupled ChIP assay was performed on the IGFBP3 promoter in MCF7 cells.
(J) MCF7 cells were transfected with either Reptin WT or K67A mutant, and immunoprecipitation was performed using anti-HDAC1 and immunoblotted using antibodies indicated.
(K) MCF7 cells with Reptin knockdown by shReptin reconstituted with exogenous shRNA-resistant Reptin WTR and K67AR mutant were exposed for indicated times in hypoxic condition and harvested for ChIP analysis examining the recruitment of various proteins on VEGF promoter as indicated.
(L) ChIP assay was performed in MCF7 cells exposed for indicated times in hypoxic condition with HDAC1 knockdown by shHDAC1.
See also Table S3.
Molecular Cell  , 71-85DOI: ( /j.molcel )
Copyright © 2010 Elsevier Inc. Terms and Conditions

7. Figure 6 Biological Consequences of Reptin Methylation
(A) Measurement of VEGF by quantitative ELISA from conditioned media collected from MCF7 cells expressing various constructs as indicated. Values are expressed as mean ± SEM of three independent experiments.
(B) Photomicrographs from a scratch-mobility assay of MCF7 cells expressing either Reptin WT or K67A in the presence of DFA (100 μM) for the indicated times.
(C) Immunoprecipitation of methylated Reptin with either anti-Reptin K67 methyl antibody or normal IgG from DFA-treated MCF7 cell lysates followed by immunoblot analysis with anti-Reptin antibody.
(D and E) Photomicrographs (40×) from transwell motility assay of MCF7 cells expressing shNS and shReptin (D) and Reptin WT or Reptin K67A (E) in normoxic and hypoxic conditions for 16 hr. Bar graph shows mean number of cells per filter, and p value is shown from Student's t test analysis. Error bars represent SEM; n = 5 (right).
(F) Matrigel invasion assay using cells expressing the indicated constructs and exposed to hypoxia for 24 hr. Bar graph shows mean number of cells per filter, and p value is shown from Student's t test analysis. Error bars represent SEM; n = 5 (right).
(G) In vivo xenograft assay of MCF7 cells expressing the indicated constructs (top). Tumors excised from nude mice (bottom) and statistical analysis of mean tumor weights; p values are shown from Student's t test analysis. Error bars represent SEM from number of tumors indicated.
(H) HIF-1α and G9a protein levels were examined from MCF7 cells expressing various constructs as indicated, as well as xenografts derived from (G) as shown. HIF-1α and G9a protein expression levels in MCF7 cells, grown in vitro in normoxia, were shown as a comparison to the levels observed in the extracted xenografts. Tubulin was used as a loading control.
(I and J) Proliferation was monitored for 3 days in hypoxic condition for MCF7 cells ectopically expressing either VEGF or FOS following shRNA-mediated knockdown of Reptin and reconstitution of either Reptin WTR or K67AR. Values are expressed as mean ± SEM of triplicate cell counts.
See also Figure S4.
Molecular Cell  , 71-85DOI: ( /j.molcel )
Copyright © 2010 Elsevier Inc. Terms and Conditions

8. Figure 7
Model for Reptin Methylation-Mediated Negative Regulation of HIF-1 Target Genes
Hypoxia induces Reptin methylation as a result of increased G9a levels. Methylated Reptin is recruited to the promoters of a subset of hypoxia target genes via enhanced binding to HIF-1α, thereby negatively regulating HIF-1 transcriptional activity.
Molecular Cell  , 71-85DOI: ( /j.molcel )
Copyright © 2010 Elsevier Inc. Terms and Conditions